库尔勒香梨内部品质及病害的声振响应多参数检测方法研究

Korla pear (Pyrus bretschneideri rehd) is a native in Xinjiang. But there exits obvious difference in its internal quality due to the increasing plant area. Also, the rough peel fruits and brown-heart as an internal defect frequently occur in pears. These problems cause heavy adverse effects on pear grade and reputation. Non-destructive evaluation of internal quality and defect is being sought by pear fruit industry to improve commodity rate. Therefore, the objectives of this research are to measure acoustic response by piezoelectric bend transducers to obtain reliable conditions for accurately determining response parameters including resonance frequency, amplitude in frequency band and sound propagation velocity in synchronization mode. Then, the effects of internal quality differences and defect development on acoustic characteristic are to be investigated. Furthermore, the mathematical relations between Young’s modulus and response parameters of pear are to be theoretically discussed. Combining with vibration modal analysis and mechanical wave propagation behavior of pear structure, the firmness evaluation indexes based on acoustic method involving resonance frequency and sound propagation velocity are to be established for pears in different shapes, respectively. Finally, the effectiveness of the considered firmness indexes and their acoustic parameters is to be tested and compared aiming to determine the optimal non-destructive evaluation method. The results can provide base in measuring internal quality and defect in real time and result in significant advantage for the commercial-scale grading line for Korla pear. Also, it has important reference value for other fruits to study non-destructive testing method of internal quality.

库尔勒香梨是新疆特色果品，由于其产地扩大所造成的内部品质差异以及粗皮果和黑心病，导致香梨等级和商品率严重下降，因此迫切需要解决其内部病害及品质差异无损检测这一关键科学问题。为此，本项目采用梁式加速度计进行香梨声振响应特性的测试分析，找到同步获取共振频率、频带幅值、声速各响应参数的稳定可靠条件，揭示香梨内部品质差异和病害程度对各响应参数的影响规律；然后在此研究基础上，对香梨弹性模量与声振响应参数之间的数学关系进行理论探讨，并结合香梨振动模态及机械波传播特性的分析，分别建立基于共振频率和声速评估不同果形香梨坚实度的指标；最后，比较分析各响应参数及相应评估指标对香梨内部病害及品质检测的精度，以获取最适宜的检测方法，为今后香梨内部品质自动化在线检测与分级装置的研制提供研究基础，同时也为其它梨果内部品质的无损检测研究提供重要借鉴。

库尔勒香梨因其内部品质差异、粗皮果以及黑心病而导致等级和商品率严重下降，迫切需要实现其硬度及内部病害的无损检测。为此，本项目采用振动模态分析方法构建了可消除果形影响的硬度评估指标，并在此研究基础上，采用压电梁式传感器测试分析香梨的声振响应特性，从时域和频域信号提取多响应特征参数，分别构建了适于香梨硬度和内部褐变的无损检测模型，为香梨内部品质在线检测与分级装置的研发提供研究基础。本研究重要结果及关键数据如下：. ⑴香梨振动的扁圆-扁椭模态和第一类球形模态分别在萼端和赤道部振动变形最大，最适宜作为信号感测位置，这两部位获取的固有频率f1、f2随着果形纵横比增大分别呈线性递减和递增，可采用这两频率建立了消除果形影响的新硬度评估指标，即双频指标：Sf1, f2 =(0.4f1+0.6f2)2m2/3ρ1/3。. ⑵采用8个频带幅值参数构建了PLSR香梨硬度检测模型，该检测模型对香梨硬度变化的敏感度为55.36%，硬度检测正确率为82.2%，对较软香梨判别精度较高，不适于较硬香梨判别。. ⑶与各单频和双频硬度指标相比，本研究提出的双频指标Sf1, f2对香梨硬度变化的敏感度为72.22%，最接近香梨硬度穿刺法测量敏感度，所构建的硬度检测模型对香梨正常果和粗皮果判别准确率最高，可达93.42%，漏判率仅为5.0%。. ⑷两种声速评估指标对香梨硬度变化的敏感度超过66%，所构建的硬度检测模型对香梨正常果和粗皮果判别准确率都超过90%，略低于本研究提出的双频指标检测模型，当香梨果形变化不大时，声速法检测装置因结构简单更适宜检测香梨硬度。. ⑸基于SVM的褐变分类器最适合采用时域信号的波形因子、方根幅值与频域信号的频率方差、均方频率所组合的特征参数集构建，对内部中度褐变香梨的判别准确率为93.9%，轻度褐变分类器最适合采用时域的波形因子、裕度因子与频域的均方频率所组合的特征参数集构建，对轻度褐变病害的判别准确率为86.4%。